The Identification of CVP1 Reveals a Role for Sterols in ...2046 The Plant Cell and the more abundant sterols and brassinosteroids (BRs) (Diener et al., 2000). SMT2 and SMT3 catalyze

The Plant Cell, Vol. 14, 2045–2058, September 2002, www.plantcell.org © 2002 American Society of Plant Biologists

The Identification of

CVP1

Reveals a Role for Sterols in Vascular Patterning

Francine M. Carland,

a

Shozo Fujioka,

b

Suguru Takatsuto,

c

Shigeo Yoshida,

b

and Timothy Nelson

a,1

a

Department of Molecular, Cellular, and Developmental Biology, Yale University, New Haven, Connecticut 06511

b

Institute of Physical and Chemical Research, Wako-shi, Saitama 351-0198, Japan

c

Department of Chemistry, Joetsu University of Education, Joetsu-shi, Niigata 943-8512, Japan

Vascular cell axialization refers to the uniform alignment of vascular strands. In the Arabidopsis

cotyledon vascularpattern1

(

cvp1

) mutant, vascular cells are not arranged in parallel files and are misshapen, suggesting that

CVP1

has arole in promoting vascular cell polarity and alignment. Characterization of an allelic series of

cvp1

mutations revealedadditional functions of

CVP1

in organ expansion and elongation. We identified

CVP1

and found that it encodes STEROLMETHYLTRANSFERASE2 (SMT2), an enzyme in the sterol biosynthetic pathway. SMT2 and the functionally redundantSMT3 act at a branch point in the pathway that mediates sterol and brassinosteroid levels. The

SMT2

gene is ex-pressed in a number of developing organs and is regulated by various hormones. As predicted from SMT2 enzymaticactivity, the precursors to brassinosteroid are increased at the expense of sterols in

cvp1

mutants, identifying a role forsterols in vascular cell polarization and axialization.

INTRODUCTION

The plant vascular system is formed progressively duringembryogenesis and postembryonic development. In organprimordia, provascular cells appear among ground cells inhierarchical patterns that give rise to the patterns of veins(Nelson and Dengler, 1997). Provascular cells proliferateand elongate in axialized files that appear to depend on theintercellular polar transport of auxin through the developingtissue. Screens for pattern-defective mutants have identifiedseveral genes that regulate or influence this process. The

monopteros

mutation causes a discontinuous venationpattern associated with improper vascular differentiation(Przemeck et al., 1996) and corresponds to an auxin re-sponse transcription factor (Hardtke and Berleth, 1998).

In

gnom/emb30

mutants, which correspond to a brefeldinA–sensitive ADP-ribosylation factor guanine-nucleotide ex-change factor (ARF-GEF), veins are unaxialized and pat-terned irregularly (Mayer et al., 1993; Shevell et al., 1994;Steinmann et al., 1999; Grebe et al., 2000; Koizumi et al.,2000). Additional venation pattern mutants have been de-scribed, such as the

van

mutants,

lop1

/

tornado1

, and

scarface

, but their gene products have not been identified(Carland and McHale, 1996; Cnops et al., 2000; Deyholos etal., 2000; Koizumi et al., 2000).

We recently described a number of additional mutants withvarious effects on cotyledon vascular pattern (Carland et al.,1999). One of these,

cotyledon vascular pattern1

(

cvp1

), re-sembles the

monopteros

phenotype in that a discontinuous,poorly axialized venation pattern is formed. Here, we showthat

cvp1

corresponds to a defect in sterol biosynthesis.Derivatives of the sterol pathway in plants and animals in-

clude both membrane sterols and signaling steroids. Re-cently, the membrane sterol cholesterol was shown to influ-ence signaling and vectorial processes through its key rolein the organization of lipid rafts (Simons and Ikonen, 1997).Lipid rafts appear to provide scaffolding for protein localiza-tion and distribution, membrane trafficking, and membranesignaling (Simons and Toomre, 2000). Although such raftshave not yet been described in plants, the diversity of plantsterols makes the possibility intriguing, particularly in highlypolarized cells such as those in procambial strands.

In contrast to animals and fungi, in which a single majorsterol is accumulated, plants accumulate many sterols(Hartmann, 1998). The most abundant plant sterols differmainly by the number of carbon additions at the C-24 posi-tion, which are catalyzed by sterol methyltransferases(SMTs) in an

S

-adenosyl Met–dependent manner (Benveniste,1986). In Arabidopsis, there are three SMTs, each of whichcan complement the defects in yeast SMT-deficient

erg6

mutants (Husselstein et al., 1996; Bouvier-Nave et al., 1997;Diener et al., 2000). SMT1, which is the most homologouswith yeast ERG6, catalyzes the initial sterol methyltransferreaction and serves as a branch point between cholesterol

1

To whom correspondence should be addressed. E-mail [email protected]; fax 203-432-5632.Article, publication date, and citation information can be found atwww.plantcell.org/cgi/doi/10.1105/tpc.003939.

2046 The Plant Cell

and the more abundant sterols and brassinosteroids (BRs)(Diener et al., 2000). SMT2 and SMT3 catalyze a secondmethyltransfer, which distinguishes the sterols from BR pre-cursors (Bouvier-Nave et al., 1997). As a result of SMT activ-ities, cholesterol, stigmasterol, campesterol, and sitosterolaccumulate (in order of increasing abundance), as do anumber of downstream derivatives.

The phenotypes of sterol biosynthetic mutants suggestthat there are diverse sterol-dependent processes. The singleSMT1 in yeast is defined by

erg6

mutants, which were recov-ered in mutant screens for impaired Trp uptake, increased re-sistance to polyene antibiotics, hypersensitivity to variousmetals and ion salts, sensitivity to brefeldin A, suppression ofvesicle trafficking, decreased mating efficiency, and modula-tion of steroid receptor signaling (Parks et al., 1999). The cau-sality between sterol deficiency and phenotype may have adifferent basis in each case and is unlikely to be limited to al-terations in membrane rigidity or permeability. Arabidopsis

smt1

mutants, in which cholesterol is aberrantly abundant,have compact rosettes, slightly reduced plant statures withblunt siliques, and a conditional root growth defect (Diener etal., 2000). Tobacco plants that overexpress soybean SMT1have decreased cholesterol and increased sitosterol and 24-methyl cholesterol without significant effects on plant growthand development (Sitbon and Jonsson, 2001).

Antisense SMT2 plants, which have lower sterol levels andhigher campesterol levels, exhibit dwarfism accompanied byreduced apical dominance, floral organ elongation, and fertility(Schaeffer et al., 2001). Plants that overexpress SMT2 containhigher levels of sitosterol and lower levels of campesterol andexhibit reduced plant stature that can be rescued with exoge-nous application of BRs (Schaeffer et al., 2001).

fackel

mu-tants, which are defective in C-14 reductase, display alteredembryonic and postembryonic patterning, including a vascu-lar patterning defect, as a result of abnormal cell division andexpansion. In

fackel

mutants, both BRs and sterols after thedefective step are reduced and uncommon 8,14-diene sterolsaccumulate (Jang et al., 2000; Schrick et al., 2000).

We cloned the gene that corresponds to the

cvp1

mutantdescribed above and found that it encodes the ArabidopsisC-24 SMT2. As predicted from the role of SMT2 in the sterolbiosynthetic pathway, sterol levels were altered significantly.Based on the phenotypes of an allelic series of

cvp1

mu-tants, we propose that

cvp1

mutants are defective in sterolsignaling or an aspect of membrane organization that is es-sential for the polarization and axialization of vascular cells.

RESULTS

Additional Alleles of

cvp1

Exhibit Defects Consistent with Aberrant Cell Expansion and Cell Elongation

As reported previously, four alleles of

cvp1

(

cvp1-1

,

cvp1-2

,

cvp1-3

, and

cvp1-4

) were identified from a screen of

�

34,000 ethyl methanesulfonate–mutagenized M2 seed-lings derived from 10 M1 pools (Carland et al., 1999).

cvp1-1

and

cvp1-2

showed identical phenotypes and were found tohave the same mutation (described below). Therefore, wecompared in detail only the phenotypes of

cvp1-1

,

cvp1-4

,and

cvp1

-

3

, which correspond to weak, intermediate, andstrong alleles, respectively.

All

cvp1

alleles exhibited a discontinuous cotyledon vena-tion pattern that consisted of isolated patches (Figures 1A to1D) of vascular tissue that lacked cellular axialization. Therewas a reduction of vein loops in cotyledons of

cvp1-4

and

cvp1-3

, and the veins were thicker. The reduced cotyledonexpansion of

cvp1-4

and

cvp1-3

in the proximal-distal axisresulted in a “soup spoon”–shaped cotyledon.

cvp1-3

coty-ledons were smaller than those of the wild type. No vascularaberrations were apparent in the rosette leaves of the

cvp1

alleles (Figures 1E to 1H). Flowers in all

cvp1

alleles werenormal in gross morphology. However, petals appearedscalloped, with highly serrated margins (Figures 1I and 1J),and sepals were highly serrated, with ectopic protrusionsextending from the apical margins (Figures 1K and 1L).These affected foliar organs often exhibited vascular alter-ations such as vein regions with nonuniform thickening.

cvp1-1

and

cvp1-4

were similar to wild-type plants ingeneral morphology and growth rate (Figure 1N), whereas

cvp1-3

plants had retarded growth and were smaller at boththe seedling and rosette stages (Figures 1M and 1N). Thecompact

cvp1-3

rosette was composed of small vegetativeleaves that failed to expand to wild-type size. This strong

cvp1-3

allele also exhibited delayed flowering time and se-nescence and reduced plant height and apical dominance.

cvp1

siliques were shorter (wild type, 13.1

�

1.1 mm;

cvp1

, 10.1

�

0.5 mm [

n

�

61 siliques from 12 plants]) andwider than wild-type siliques (Figure 1O). However, a reduc-tion in fertility occurred only in the strong

cvp1-3

allele. Asdescribed previously,

cvp1-1

had a decrease in internodalstem elongation in the apex of primary and secondary inflo-rescences in

�

10% of the inflorescence stems. Transversesections of stem tissue within this mutant region revealed anoverproliferation of lignified tissue (Carland et al., 1999). Thisphenotype occurred at a higher frequency in

cvp1-3

, affect-ing 55% of plants (77 of 139) (Figure 1P). In summary, in ad-dition to the loss of vascular cell axialization, strong

cvp1

mutants exhibited defects in elongation at the cellular level(vascular cells) and/or the organ level (stem and siliques)and an expansion defect similar to that seen in leaves.

CVP1

Encodes SMT2

We previously mapped

CVP1

to chromosome 1 near

MONOPTEROS

between m59 and m235 (Hardtke and Berleth,1998). To clone

CVP1

, a large mapping population was gen-erated and scored for recombinants, as diagrammed in Fig-ure 2A. Polymorphic markers were designed based on lim-ited sequence data in the region and used to further localize

Sterol Modification Alters Vein Patterns 2047

CVP1

to two overlapping BACs. PCR products were ampli-fied from the unordered BAC sequences and used asprobes to screen a transformation-competent artificial chro-mosome (TAC) library (Liu et al., 1999), resulting in an or-dered contig of TAC clones spanning the

CVP1

region.Individual TAC clones were introduced into

cvp1-1

mu-tants and scored for complementation of the

cvp1

mutantphenotype. A complementing TAC contained six uniquegenes relative to overlapping noncomplementing TACs,based on gene annotation that had become available bythat date. These six genes were sequenced in the

cvp1-1

background, and a mutation was identified in a previouslyidentified gene encoding SMT2 (Bouvier-Nave et al., 1997).The additional

cvp1

alleles contained mutations in this gene.

cvp1

was complemented by transformation with the

SMT2

genomic region, confirming the identity of

CVP1

(Figure 2B).Although

cvp1-1

and

cvp1-2

were identified from differentM1 pools and therefore should be independent alleles, theycontain the same mutation in an invariant residue within thefirst

S

-adenosyl Met binding site.

cvp1-4

was found to havea nonsense mutation that is predicted to produce a trun-cated protein. The

cvp1-3

allele contained a missense

Figure 1. Phenotypes of cvp1 Alleles.

(A) to (H) Venation patterns of cleared cotyledons (A) to (D) and first rosette leaves (E) to (H) viewed under dark-field illumination. Arrows denotevascular discontinuities.(A) and (E) Wild type (Columbia ecotype).(B) and (F) cvp1-1.(C) and (G) cvp1-4.(D) and (H) cvp1-3.(I) to (L) Margins of cleared petals (I) and (J) and sepals (K) and (L) viewed under dark-field illumination. Arrows indicate smooth margins of thewild type in contrast to the marginal protrusions of cvp1-1.(I) and (K) Wild type.(J) and (L) cvp1-1.(M) Seven-day-old seedlings of the wild type (WT) and cvp1-3. The growth of cvp1-3 is retarded, and seedlings are smaller than those of thewild type.(N) Vegetative stage of wild-type, cvp1-3, and cvp1-4 plants. cvp1-4 is indistinguishable from the wild type, but cvp1-3 has a compact rosette.(O) Siliques of the wild type and cvp1. cvp1 mutant siliques are shorter and wider than wild-type siliques.(P) Inflorescence stem internodal elongation defect in cvp1-3 compared with that of the wild type. Nodes between siliques fail to elongate, re-sulting in a cluster of siliques.Bars in (A) to (D) and (I) to (L) � 250 �m; bars in (E) to (H) � 500 �m; bar in (M) � 1 mm; bar in (N) � 1.5 cm; bar in (O) � 3 mm; bar in (P) � 1 cm.

2048 The Plant Cell

mutation in the initiation codon (Figure 2C) that should resultin a null allele, consistent with its more severe phenotype.The next in-frame Met residue follows the sterol binding site.

Features of the SMT2 Protein

SMT2 is one of three SMT enzymes in Arabidopsis, all ofwhich share regions of homology. SMT1 has substrate affin-ities that differ from those of SMT2 and SMT3 and acts ear-lier in the sterol biosynthetic pathway (Figure 3). SMT2 andSMT3 are 83% identical in amino acid sequence and act onthe same substrate at the pathway bifurcation that distin-guishes sterols from BRs. The predicted protein structure ofSMT2 includes a hydrophobic region at the N terminus (Fig-ure 2C). This region meets the criteria for a signal anchor (aspredicted with SignalP [http://www.cbs.dtu.dk/services/SignalP]) and may serve to direct SMT2 to the endoplasmicreticulum, where sterol biosynthesis is postulated to occur.The sequence predicts three sites for binding of S-adenosylMet, the methyl donor for SMT reactions, a sterol bindingsite (Nes et al., 1999), and a region of unknown functionconserved among all SMTs.

SMT2 Is Expressed in Several Developing Organs

We examined the effects of the cvp1 molecular lesions onSMT2 transcript levels by RNA gel blot analysis. RNA iso-lated from 6-day-old seedlings revealed decreases in SMT2transcript levels in all alleles (Figure 4A) in proportion to theirphenotypic severity, even though the mutations were pre-dicted to alter translational products rather than transcrip-tion. There were undetectable levels of SMT2 transcripts inthe severe cvp1-3 allele, confirming the designation of cvp1-3as a null allele.

The cvp1 mutations in SMT2 also affected the accumula-tion of transcripts for SMT1 and SMT3. The transcript levelsof the highly homologous SMT3 and of SMT1, determinedwith gene-specific probes, were increased in cvp1-4 but

Figure 2. Positional Cloning of CVP1.

(A) Line drawing of CVP1 cloning strategy. cvp1 in the Columbiabackground was crossed to the polymorphic ecotype Landsbergerecta to generate a mapping population of 480 meiotic events. Us-ing breakpoint analysis, CVP1 was positioned on chromosome 1 be-tween cleaved amplified polymorphic markers m59 and g2395. Thenumbers of recombinants between CVP1 and polymorphic flankingmarkers are shown below the line with double arrows. The BACsfrom which markers were designed are noted at top. CVP1 was lo-calized to the two overlapping BACs, T20H2 and F14O10. Probesdesignated by small thick lines were generated from these BACsand used to identify overlapping TAC clones as labeled. Only theTAC clones that yielded transgenic plants are shown.(B) Complementation of cvp1. One TAC clone complemented thecvp1-1 mutant vascular patterning defect (K3J15) and contained sixunique annotated genes relative to a noncomplementing TAC clone

(K3L10). A genomic clone of SMT2 was generated and shown tocomplement cvp1. Bars � 250 �m.(C) Gene structure with mutations. Motifs proposed to be function-ally important, as listed from amino acid 1, are as follows: thehatched box represents the signal anchor; the box with diagonal andhorizontal lines indicates the sterol binding site (Nes et al., 1999); theblack box denotes a conserved region among SMTs of unknownfunction; S-adenosyl Met (SAM) sites are labeled boxes. Mutationswithin the cvp1 alleles are denoted. Specifically, cvp1-3 has a G-to-Anucleotide substitution in the initiation codon, changing Met to Ile.cvp1-1 and cvp1-2 are independent alleles that have a G-to-A nu-cleotide substitution, changing Asn to Asp. cvp1-4 has a C-to-T nu-cleotide substitution, resulting in an early termination codon.

Sterol Modification Alters Vein Patterns 2049

were unaffected in cvp1-1. Remarkably, neither transcriptwas detected in cvp1-3, suggesting that the null allele is de-ficient in all SMT activity. We reproduced these measure-ments several times, because the results were unexpected.Some effects on SMT2 and SMT3 transcript levels were re-ported previously for smt1 mutants, but these were not asgreat as the effects we observed in cvp1 alleles (Diener etal., 2000). The increased SMT1 and SMT3 transcript levelsin cvp1-4 and the decreases in cvp1-3 may reflect a feed-back mechanism between sterol production and activation ofthe SMT genes. Alleles that differ in severity may disrupt do-mains in SMT2 involved in distinct aspects of this regulation.

The SMT2 gene was expressed in most developing tis-sues in a pattern that differed slightly from that of SMT3 andSMT1. RNA gel blot analysis showed that SMT2 mRNA waspresent in all developing tissues surveyed and revealed low-level expression in roots and siliques, moderate levels inseedling and rosette leaves, and abundant levels in stemsand inflorescences (Figure 4A). SMT3 showed a similar ex-pression profile, except that mRNA was at undetectable lev-els in siliques (Figure 4B). Likewise, the pattern of SMT1mRNA (data not shown) resembled that of SMT2, in agree-ment with a previous report (Diener et al., 2000).

The expression of SMT2 was influenced by several hor-mones. We measured SMT2 and SMT3 mRNA after treat-ment with ethylene, cytokinin, indoleacetic acid, epibrassin-olide, or gibberellic acid, as described in Methods. Hormoneinduction studies indicated that SMT2 was induced stronglyand moderately by ethylene and cytokinin, respectively.SMT3 was induced strongly by cytokinin. SMT2 and SMT3were induced weakly and moderately, respectively, by in-doleacetic acid. The BR epibrassinolide and gibberellic aciddid not appear to have any effect on SMT transcript levels.This finding suggests that SMT2 and SMT3 are induced dif-ferentially by at least three hormones that promote cell divi-sion and/or elongation.

SMT2 appeared to be expressed predominantly in re-gions with active cell division. We visualized SMT2 mRNAaccumulation by in situ RNA hybridization (Figure 5). As re-ported previously (Diener et al., 2000), SMT2 was expressedthroughout the developing embryo (Figure 5A). In agreementwith the RNA gel blot studies, SMT2 was expressed in anumber of nascent organs. These included the primordia ofleaves, sepals, stamens, gynoecia (Figures 5B to 5E), andpetals (data not shown). SMT2 also was expressed in the in-florescence meristem (Figure 5C) and in developing ovules(Figure 5F). SMT2 expression at these sites was limited ineach case to stages that included actively dividing cells.

�-Glucuronidase (GUS) reporter analysis revealed someadditional features of the SMT2 and SMT3 expression pat-terns. Both were expressed specifically in the elongationzone of the roots (Figures 6A and 6B). In rosettes, GUS his-tochemical staining was apparent in cotyledons, expandingleaves, and the apical tips of young leaves (Figures 6C and6D). These are regions in the leaf that are undergoing cellexpansion and reflect the basipetal maturation of leaves. In

Figure 3. The Sterol Biosynthetic Pathway.

A simplified version of the sterol biosynthetic pathway is shown.Sterol intermediates are labeled. Genes that are enzymes and thathave been identified by mutation, with the exception of SMT3, areshown in uppercase letters (Klahre et al., 1998; Choe et al., 1999a,1999b, 2000; Diener et al., 2000; Jang et al., 2000; Schrick et al.,2000). The site of SMT methyl addition, C-24, is indicated. SMT2 isshown in boldface and underlined. Brz is a BR biosynthetic inhibitor(Asami and Yoshida, 1999). Dashed arrows indicate multiple biosyn-thetic steps.

2050 The Plant Cell

SMT2::GUS and SMT3::GUS flowers, staining was localizedto the apical region of the sepal, particularly in the vascularbundles (Figures 6E and 6F). In SMT2::GUS siliques, GUSstaining also was detected in ovules (Figure 6E). These data,in conjunction with the RNA gel blot analysis of organs andhormone application, indicate that SMT2 and SMT3 havesimilar but not identical expression patterns.

SMT2 and SMT3 Are Functionally Redundant

SMT2 and SMT3 are highly homologous, act on the samesubstrate, and exhibit similar but distinguishable expressionpatterns (Figures 4 and 6), suggesting that the proteins arefunctionally redundant. Because SMT3 expression is re-stricted to the outer cell layers of the developing embryo,and thus is excluded from the majority of procambial cells ofdeveloping cotyledons (Diener et al., 2000), it appears thatthe function is provided solely by SMT2 in these locations.This probably explains the aberrant venation of cvp1 mu-tants and provided an experimental system to test for func-tional redundancy in vivo.

We expressed either SMT3 or SMT2 in the cvp1-1 back-ground, under the control of the 35S promoter of Cauliflowermosaic virus, and observed full restoration of the vascularpatterning, petal and sepal margin, and silique elongationdefects in cvp1-1 (Figures 7A to 7F and data not shown).However, the stem elongation defect was not rescued bySMT3 (Figure 7G). This finding appears to reveal a role forSMT2 in the stem that is not shared by SMT3, indicatingthat the functional overlap is extensive but not complete.

SMT2 and SMT3 have overlapping functions but are notcompletely functionally redundant. The overexpression ofSMT2 and SMT3 yielded high transcript levels in the cvp1-1background but did not produce any phenotypic abnormali-ties (Figures 7H and 7I). In addition, SMT2 overexpressiondid not have a detectable effect on SMT3 transcript levels,and likewise, SMT3 overexpression did not have an effecton SMT2 transcript levels, indicating that the cvp1 comple-mentation is attributable specifically to SMT3 ectopic ex-pression. The overexpression of SMT3 in the sense and an-tisense orientations in the wild-type background in �60transgenic plants also did not produce a phenotype (datanot shown).

Based on the SMT3 complementation of cvp1 and the ob-servation that SMT3 is not expressed independently fromSMT2, we propose that the lack of a phenotype in SMT3 an-tisense plants is caused by the functional redundancy of

Figure 4. RNA Gel Blot Analysis of SMT.

(A) SMT transcript levels in different alleles (SMT2, SMT3, andSMT1) were hybridized sequentially to filters of the cvp1 alleles sub-sequent to probe removal. RNA was isolated from 6-day-old seed-lings. WT, wild type.(B) SMT2 and SMT3 exhibit similar expression profiles. RNA wasisolated from the indicated organ. Roots were excised from 1-week-oldseedlings. Seedlings were 1 week old. Young leaves and expandingleaves were 2 to 3 mm and 4 to 5 mm in length, respectively. Si-liques were from all developmental stages. Elongating regions of thestem were selected for RNA analysis.(C) Tissue was incubated with the indicated hormone (see Methods).Mock, no hormone, BA, benzyladenine; ACC, 1-aminocyclopro-

pane-1-carboxylic acid; GA, gibberellic acid; IAA, indoleacetic acid,BL, epibrassinolide. Ten micrograms of total RNA was loaded ontolanes. Methylene blue staining of rRNA served as the loading con-trol. Only the top rRNA band is shown.

Sterol Modification Alters Vein Patterns 2051

SMT2 and SMT3. This redundancy also may provide an ex-planation for the wild-type gross morphology of cvp1-4.cvp1-4 has only 35% of SMT2 wild-type RNA levels and ispredicted to have a truncated SMT2 protein. However, in-creased levels of SMT3 mRNA accumulate in this allele. Wepropose that the wild-type morphology of cvp1-4 is main-tained by the increase in SMT3 activity, which compensatesfor the decrease in SMT2 activity. The overlap in SMT2 andSMT3 expression patterns does not include the majority of

cotyledon procambial cells, so a vascular patterning defectremains (Diener et al., 2000).

cvp1 Mutants Alter the Balance between Sterolsand BRs

Because SMT2 catalyzes the methylation that distinguishessterols from BRs, we determined whether the levels of steroland BR intermediates were affected in the cvp1 alleles. Thelevels of 22 sterol compounds were assayed in 1-week-oldcvp1-1, cvp1-3, and cvp1-4 seedlings (Table 1). Becausewhole seedlings were extracted for measurements, it is pos-sible that individual cells and tissues exhibited significantlydifferent values. In all tested alleles, the levels of intermedi-ates before 24-methylenelophenol, the substrate of SMT2,were not altered significantly. By contrast, 24-methylene-lophenol was increased in the cvp1 mutants, and productsof the sterol-specific pathway, such as 24-ethylidenelophe-nol, sitosterol, stigmasterol, and sitostanol, were decreased.24-Methylenelophenol also was channeled into the BRpathway.

These precursors to the BRs were increased in cvp1 rela-tive to the wild type. For example, the null allele cvp1-3showed more than a 25-fold increase in 24-methylenecho-lesterol. Cholesterol lacks methyl additions and is formed bybypassing the methyl addition reactions. As predicted, cho-lesterol was higher in cvp1 mutant alleles. Although therewere gross changes in individual sterol levels, the total ste-rol content of cvp1 was not markedly different from that ofthe wild type. Therefore, sterol levels were altered signifi-cantly in cvp1, most notably in the null allele cvp1-3, byshifting the balance of intermediates in favor of the BR path-way at the expense of the sterol pathway.

Because SMT2 regulates the biosynthetic flux that con-trols the amount of sterol and campesterol, the alteration inthe flux often is expressed as a ratio of campesterol to sito-sterol. In the wild type, this ratio was equal to 0.2, whereasin cvp1, the ratio was 1.9 or greater. In general, the de-crease in sterol levels of the three alleles was proportional tothe severity of their phenotypes. Although the cvp1-3 null al-lele exhibited a reduction in transcript levels in all SMTs, itwas able to produce alkylated sterols, albeit at reduced lev-els. The presence of sterols was reported in a null smt1 al-lele (Diener et al., 2000). It is possible that lateral sterol path-ways not normally used in wild-type sterol production areused in smt1 and cvp1 mutants.

The BR-Specific Inhibitor Brassinazole Does Not Rescue the cvp1 Mutant Phenotype

All cvp1 mutant alleles exhibited a decrease in sterols and aconcomitant increase in the precursors to the BRs. AlthoughBR levels were not assayed because of the large amount of

Figure 5. In Situ Localization of SMT2 mRNA.

(A) SMT2 is expressed throughout the developing embryo. Seedcoat staining is background.(B) Longitudinal section of a 7-day-old seedling to show SMT2 ex-pression in leaf primordia. The shoot apical meristem is not visible inthis section.(C) Longitudinal section of an inflorescence shows SMT2 expressionin the inflorescence meristem and sepals.(D) Slightly oblique section of a stage 3 flower showing intenseSMT2 expression in the sepals.(E) Longitudinal section of a stage 5 to 6 flower reveals SMT2 ex-pression in stamens and in the gynoecium before fusion.(F) Ovule expression of SMT2.In (C) and (E) images were photographed with a blue filter. cpt, cot-yledon petiole; gy, gynoecium; IM, inflorescence meristem; lp, leafprimordia; ov, ovules; se, sepals; st, stamens. Bars � 40 �m.

2052 The Plant Cell

tissue required, the cvp1 sterol profile suggested that therewas an increase in BR levels. Therefore, we determinedwhether the inhibition of the BR biosynthetic pathway couldrescue any aspects of the cvp1 phenotype. The cvp1 vascu-lar patterning defect occurs during embryogenesis. Thus,we applied the BR biosynthetic inhibitor brassinazole (Brz;Figure 3) (Asami and Yoshida, 1999) to soil-grown plantsbefore bolting to test its effects on the stem internodal elon-gation defect.

As shown in Figure 8A, Brz application did not rescue thisaspect of the mutant phenotype. Subsequent to treatment,seeds were isolated and the cotyledon vascular patterningdefect was assessed in seedlings. Brz application duringembryogenesis did not rescue the cvp1 vascular patterningdefect, nor did it affect the wild-type venation pattern (Fig-

ures 8B and 8C), suggesting that the phenotype is not theresult of altered BR levels. The cvp1 alleles exhibited a sig-nificant reduction in sitosterol and stigmasterol levels. Wewere unsuccessful in altering cvp1 by supplementation withthese compounds, although supplementation is made diffi-cult by the limited solubility of the compounds.

DISCUSSION

The first recognizable event associated with the differentia-tion of provascular tissue is cell polarization and the forma-tion of axialized procambial files. Our analysis of mutationsthat disrupt the SMT2 gene demonstrates that this processis dependent on a functional sterol pathway. Arabidopsishas two very similar proteins (SMT2 and SMT3) acting at thesame step in sterol biosynthesis, but only SMT2 is ex-pressed in cotyledon vascular cells. This was important fortwo reasons. First, it allowed the critical observation of aber-rant venation in the cvp1 mutants, establishing a vascularrole for the pathway. Second, it provided an experimentalopportunity to introduce ectopic SMT3, demonstratingfunctional redundancy.

SMT2 acts at the bifurcation of the sterol and BR path-ways, at which the substrate 24-methylenelophenol is meth-ylated to enter the sterol pathway or is not further methy-lated and enters the BR pathway. Thus, SMT2 regulates theratio of campesterol, precursor of the BRs, to sitosterol, themajor sterol in plants. In cvp1 mutants, we showed that thisratio, which normally favors the sterol pathway, was invertedto favor the BR pathway, possibly most dramatically in coty-ledons. The phenotypic consequences could be the resultof the enrichment of one or more BRs or of the depletion ofone or more sterols, or a combination of both effects.

We favor the second of these possibilities. We believethat the potential increase of BR levels is unlikely to be thecause of the cvp1 phenotype, for several reasons. First, wewere unable to phenocopy the cotyledon vascular defect bysupplementation of wild-type plants with brassinolide. Sec-ond, we were unable to cure the cvp1 mutant phenotype bythe application of Brz, an inhibitor that acts at a point in theBR pathway downstream from campesterol. Third, the BRpathway appears to be highly regulated, such that increasesof early intermediates do not necessarily result in increasedlevels of the downstream active BR, brassinolide (Clouseand Sasse, 1998). Although we cannot exclude the possibil-ity that BR alterations cause some of the cvp1 phenotype,we believe it more likely that sterol depletions or modifica-tions are responsible.

Sterols as membrane-ordering components are likely tohave roles in organizing the many asymmetric structuresand processes that are essential for the development of axi-alized files of polarized cells, such as the provascular cellsand procambial strands that are affected in cvp1 mutants. Inaddition, the sterol branch of the pathway may produce

Figure 6. GUS Expression Studies of SMT2 and SMT3.

(A), (C), and (E) SMT2::GUS reporter expression.(B), (D), and (F) SMT3::GUS reporter expression.(A) and (B) Two-day-old germinated seedlings. Note the intenseGUS staining in the elongation zone of the root.(C) and (D) Plants with rosette leaves. In young leaves, GUS activityis restricted to the apical zone (arrows).(E) and (F) Detached flowers show GUS enzyme activity in sepals,veins, and ovules ([E] only). For the inset in (E), outer floral organshave been removed to expose the carpel.ez, elongation zone; ov, ovules. Bars in (A) and (B) � 250 �m; barsin (C) to (F) � 2 mm.

Sterol Modification Alters Vein Patterns 2053

compounds distinct from BRs, with roles in intercellular sig-naling for the alignment of cell files and elongation axes or inconferring cell identity distinct from neighbor cells. We dis-cuss each of these processes below in relation to the cvp1phenotype.

Role of SMT2 in Signaling

The cvp1 venation defect may be the result of the absenceor alteration of a sterol not identified previously as an inter-cellular or intracellular signal. Such a signal might serve toalign provascular cells along an axis or to prevent neighbor-ing cells from joining in the provascular cell fate. Vascularpatterning is altered in several biosynthetic mutants that af-fect both sterols and BRs, including fackel, which is blockedupstream of the sterol/BR bifurcation, and dwf7/ste1 (Choeet al., 1999b), which is blocked downstream. In addition,bri1 mutants, with a defective BR receptor, also exhibit ab-errant venation (Li and Chory, 1997; Wang et al., 2001). Un-like cvp1, these mutants generally show an overproliferationof phloem at the expense of xylem (Choe et al., 1999a).However, weak fackel alleles exhibit discontinuous venationpatterns in cotyledons, similar to cvp1 (Jang et al., 2000).

Although we did not measure brassinolide, the active sig-nal in Arabidopsis, directly, it is unlikely that the cvp1 defectis caused by excessive BR levels, for the reasons describedabove. Instead, another sterol-derived signal may be af-fected in cvp1 and fackel mutants. There is indirect evi-dence for such signaling sterols in the spatial organization ofvarious organs. Sterol/lipid binding domains (START) havebeen identified in a subtype of homeodomain/Leu zipper–containing proteins, which includes Athb-8, Athb-9, Athb-14, and REVOLUTA/INTERFASCICULAR FIBERLESS (Sessaet al., 1998; Zhong and Ye, 1999; Ratcliffe et al., 2000). Inthe case of the radial patterning genes Phabulosa (phab)and Phavoluta (phav), which encode the proteins ATHB-14and ATHB-9, respectively, an unidentified sterol or lipid withspatial distribution is postulated to bind to the START do-main of the receptors PHAB and PHAV to activate the genesinvolved in radial patterning (McConnell et al., 2001).

SMT2 and the Synthesis of Cellulose Microfibrils

The cvp1 sitosterol deficiency might cause a vein pattern de-fect by interfering with the pattern of cellulose synthesis in the

Figure 7. Functional Redundancy of SMT2 and SMT3.

(A) to (F) Cleared images of cotyledons ([A] to [D]) and petals ([E]and [F]) viewed under dark-field illumination. Bars � 250 �m.(A) Wild type.(B) cvp1-1.(C) and (D) cvp1-1 transformed with 35S/SMT2 (C) and 35S/SMT3(D) to show complementation of the vascular patterning defect.(E) and (F) cvp1-1 petal with a scalloped apical region (E) and cvp1-1transformed with 35S/SMT3 (F) to show complementation of thepetal defect.(G) The stem internodal elongation defect is not rescued by SMT3overexpression. A normal cvp1-1 inflorescence stem is shown atleft; an affected inflorescence stem of cvp1-1::35S/SMT3 is shownat right. Three of 36 plants had affected inflorescences. Bar � 5 mm.(H) and (I) RNA gel blot analysis of SMT2 (H) or SMT3 (I) overex-pressors (OE). The first lane contains RNA from wild-type (WT) seed-lings. The transcripts are weak as a result of a short exposure time.The next three lanes contain RNA isolated from three different SMT2

or SMT3 overexpressors. Each lane contained 5 �g of total RNA iso-lated from 9-day-old seedlings.

2054 The Plant Cell

walls of provascular cells. Recently, the major plant sterol sito-sterol was suggested as a primer for cellulose synthesis, whichis initiated with the conjugation of Glc to sitosterol, formingsitosterol-�-glucoside (Peng et al., 2002). Further addition ofGlc residues forms sitosterol cellodextrin, which is cleaved, al-lowing the transfer of the cellodextrin chain to another cellu-lase. After the attachment of additional Glc residues, the chainthen assembles into cellulose microfibrils, which are the pri-mary scaffolding component of plant cell walls and are postu-lated to orient the direction of cell expansion.

Because sitosterol is reduced greatly in cvp1, both sito-sterol-�-glucoside and cellulose microfibrils might be re-duced as a consequence, causing aberrations in the patternof cell wall extensibility that normally guide the elongationand morphology of provascular and vascular cells. The ab-errant cell expansion patterns observed in cvp1 foliar organsand in fackel may be attributable to the same effects on cel-lulose priming. As discussed below, cobra mutants have re-duced levels of cellulose and a concomitant alteration in cel-lulose microfibrils and cell expansion (Schindelman et al.,2001).

Lipid Rafts and the Compartmentalization of Proteins

In other systems, sterols have been shown to influence pro-tein activity through their central role in organizing lipid rafts.Lipid rafts are membrane microenvironments that consist ofdynamic compositions of sterols and lipids; they were iden-tified initially in polarized epithelial and neuronal cells, inwhich they accumulate at a polar locality in the plasmamembrane (Simons and Ikonen, 1997; Simons and Toomre,2000). Through the inclusion or exclusion of specific pro-teins on the basis of affinity, rafts lead to the clustering ofmembrane-bound and membrane-associated proteins andserve as compartments for protein signaling complexes, cy-toskeletal associations, channels, and other functional featuresof membranes. Numerous proteins, such as glucosylphos-phatidylinositol (GPI)-linked proteins and transmembraneproteins, have been demonstrated to reside in such rafts. Inmany cases, sterol depletion results in the dissociation ofmembrane-bound proteins from lipid rafts, leading to mislo-calization of the protein and/or inactivation of the signalingcascade (Simons and Toomre, 2000).

Rafts are assembled initially with endoplasmic reticulum–derived sterols in the Golgi, the site of lipid synthesis, andthen are transported to the plasma membrane carrying pro-teins destined for polarized delivery. After protein delivery,rafts are endocytosed from the cell surface and then recy-cled either directly back to the plasma membrane or indi-rectly by endosomes. In plants, there are low sterol levels inthe endoplasmic reticulum, where sterols are synthesizedas corroborated by the presence of signal anchors onSMT2 and SMT3, and high levels in the plasma membrane,providing support for this mode of transport.

Although lipid rafts have not been observed in plants,their small size may have precluded observation by conven-tional microscopy. In other systems, raft components havebeen resolved through advanced microscopy techniques cou-pled with biochemical methods. Recently, a GPI-anchoredprotein was found to be the product of the ArabidopsisCOBRA gene, whose mutant phenotype includes a loss ofpolarized, longitudinal expansion (Schindelman et al., 2001).The COBRA protein is localized asymmetrically in polarizedcells, much like GPI-anchored proteins in other systems,and is postulated to play a role in oriented cell expansion,possibly through the recruitment of enzyme components in-volved in cellulose deposition. Although not demonstrateddirectly, this localization has the properties expected of amembrane raft-localized protein.

Other candidates for lipid raft localization in plants are thePIN family of auxin efflux carriers (Galweiler et al., 1998).Auxin flow is postulated to direct vascular cell patterningthrough the vectorial transport of auxin by basally localizedefflux carriers. The mislocalization of a PIN efflux carrier incotyledons of cvp1 mutants might lead to the failure to formaxialized files of provascular cells. PIN is cycled betweenthe plasma membrane and an undefined endosomal com-partment requiring ARF-GEF (Geldner et al., 2001) encoded

Table 1. Sterol Profile of cvp1 Alleles

Genotype

Sterol Wild Type cvp1-1 cvp1-4 cvp1-3

Cycloartenol 0.40 0.38 0.25 0.2924-Methylenecycloartanol 0.49 0.45 0.25 0.29Cycloeucalenol 0.26 0.34 0.22 0.29Obtusifoliol 0.56 0.58 0.22 0.304�-Methyl-5�-ergosta-8,

14,24(28)-trien-3�-ol 0.03 0.03 0.02 0.03

24-Methylenelophenol 0.06 0.87 0.79 0.8924-Ethylidenelophenol 0.35 0.13 0.10 0.12Avenasterol 0.24 0.10 0.10 0.11Isofucosterol 1.8 1.3 1.5 1.9Sitosterol 91 39 32 26Stigmasterol 8.8 3.9 5.0 4.5Sitostanol 1.14 0.36 0.63 0.336-Oxositostanol 0.13 0.06 0.06 0.03Episterol 0.053 0.42 0.38 0.6324-Methylenecholesterol 0.27 2.90 2.62 6.82Campesterol 19.2 72.3 67.5 68.9Campestanol 0.33 1.01 1.31 1.306-Oxocampestanol 0.022 0.073 0.040 0.0916-Deoxocathasterone 0.0031 0.0041 0.0031 0.0033Cholesterol 1.9 4.2 3.4 6.7Cholestanol 0.08 0.09 0.13 0.266-Oxocholestanol 0.02 0.03 0.02 0.03Total sterol content 127.0 128.5 116.5 120.0Ratio of campesterol to

sitosterola0.21 1.9 2.2 2.6

Values are given in �g/g fresh weight.a Values were converted to fraction of sterol/total sterol before calcu-lating the ratio.

Sterol Modification Alters Vein Patterns 2055

by EMB30/GNOM (Shevell et al., 1994; Busch et al., 1996;Steinmann et al., 1999; Geldner et al., 2001). Although PIN-containing vesicles are not derived directly from the Golgi,such cycling is seen with other lipid raft-associated proteins.

It is tempting to speculate that PIN asymmetric deliveryacts in association with lipid rafts. Alternatively, PIN may re-side in plasma membrane microdomains rich in sterols andlipids. Although pin mutants have not been reported to ex-hibit any cotyledon vascular patterning discontinuities,

emb30/gnom mutants exhibit a loss of vascular cell axializa-tion similar to that observed in cvp1 mutants, albeit muchmore severely (Mayer et al., 1993; Koizumi et al., 2000).There are many members of the PIN family, so functional re-dundancy probably exists. In emb30/gnom, vesicle transport–dependent polar localization of all PIN members would beaffected; thus, a much more severe phenotype would occur.

Based on the known role of sterols in mammalian andyeast cells, it is likely that sterols act in more than one pro-cess. We propose that the cvp1 mutant phenotype is notcaused by a disruption in membrane permeability but ratherreflects either (1) the mistargeting of proteins involved in po-larizing and aligning vascular cells or (2) the synthesis ofnovel sterol signaling molecules. These two possibilities arenot mutually exclusive. Sterol signaling may involve a recep-tor that subsequently initiates a signaling cascade that re-sides in a plasma membrane microdomain. The further anal-ysis of cvp1 and other sterol-defective mutants shouldreveal the roles of sterols in vascular pattern formation andother developmental processes.

METHODS

Plant Growth Conditions

All cvp1 alleles were recovered from an ethyl methanesulfonate–mutagenized population of Arabidopsis thaliana ecotype Columbia inthe screen described previously (Carland et al., 1999) and back-crossed at least three times. Seeds were surface-sterilized, plated onMurashige and Skoog (1962) (MS) medium (Sigma) with (for trans-genic plants) or without antibiotics, and grown under continuouslight. Approximately 1-week-old seedlings were either collected forRNA gel blot analysis, �-glucuronidase (GUS) staining, or sterol anal-ysis or transplanted to soil and grown to maturity for additional ma-nipulations. Wild-type and mutant plants were at the same develop-mental stage for comparative morphological studies. Plants in soil(Metromix; Scotts, Hope, AR) were grown at 22�C with a 16-h day anda light intensity of 200 �mol·m2·s1. The phenotype of cvp1-3 plantsgrown at a higher light intensity (290 �mol·m2·s1) was not as severe.

Histology and GUS Histochemical Staining

To score cotyledon defects, one cotyledon was detached andcleared of chlorophyll by immersing cotyledon in fixative (etha-nol:acetic acid [3:1]) and then rinsed in 70% and incubated in 100%ethanol at 4�C overnight. Cotyledons were viewed with a dissectingmicroscope. For photography, after the 100% ethanol dehydrationstep described above, specimens were further cleared by 1 h of in-cubation in 10% NaOH at 42�C and mounted in 50% glycerol. Vas-cular patterns were viewed under dark-field illumination using a ZeissAxiophot microscope (Jena, Germany).

Plant organs were placed in 5-bromo-4-chloro-3-indolyl-�-gluc-uronic acid (X-Gluc) staining solution, infiltrated under vacuum for 10min, wrapped in foil, and placed at 37�C. GUS staining was monitoredfor 2 to 12 h. X-Gluc staining solution of aerial organs contained

Figure 8. The BR Inhibitor Brz Does Not Rescue the cvp1 MutantPhenotype.

(A) Failure of Brz to rescue the stem elongation defect of cvp1-3.Note that there is no increase in plant height in Brz-treated cvp1plants relative to untreated cvp1 plants.(B) and (C). Cleared cotyledons of wild-type (B) or cvp1-3 (C) seed-lings whose embryos were treated with Brz, viewed under dark-fieldillumination.Plants were watered with 1 �M BRZ2001 (Sekimata et al., 2001)before bolting and until dessication. For (B) and (C) seed was col-lected from Brz-treated plants and germinated on Murashige andSkoog (1962) medium without Brz. Bar in (A) � 2.5 cm; bars in (B)and (C) � 100 �m.

2056 The Plant Cell

1 mM X-Gluc, 1 mM Na-phosphate buffer, pH 7.0, 1.0 mM EDTA, pH8.0, 0.1% Triton X-100, 1 mM �-mercaptoethanol, 100 �g/mLchloramphenicol, 0.5 mM K3Fe(CN)6, and 0.5 mM K4Fe(CN)6. Im-proved specificity of X-Gluc staining in roots was obtained by follow-ing the method of Malamy and Benfey (1997). Samples were rinsedin 70% ethanol two times and mounted in 50% glycerol for photog-raphy using a Zeiss Stemi 2000-C dissecting microscope.

Positional Cloning of CVP1

cvp1-1 (ecotype Columbia) was crossed to Landsberg erecta. Theresulting hybrid was self-pollinated to generate a large mapping pop-ulation. Medium-grown F2 seedlings were scored for the mutantcotyledon vascular patterning phenotype. cvp1 mutants were trans-ferred subsequently to soil and allowed to mature. DNA was isolatedfrom leaf tissue and scored for recombinants with the flankingcleaved amplified polymorphic markers g2395 and m59 (Hardtkeand Berleth, 1998). Seeds were harvested from the cvp1/g2395 andcvp1/m59 recombinants. At this time, we noticed that cvp1-1 in theLandsberg erecta background had a pleiotropic phenotype that in-cluded dwarfism and a severely stunted silique growth defect ac-companied by a drastic reduction in fertility. This phenotype did notsegregate with the Landsberg erecta mutation.

CVP1 was localized further by searching for polymorphic markers.To obtain new polymorphic markers, as BAC sequences becameavailable in this region, 0.8-kb regions of DNA were amplified by PCRfrom Landsberg erecta, sequenced, and compared with the Colum-bia sequence. We designed several novel markers with this methodand used them to further define the position of CVP1 to an �150-kbregion spanning two BACs. These markers are available upon re-quest. Transformation-competent artificial chromosome (TAC) filters,obtained from the ABRC (Columbus, OH), were probed with 1-kbprobes amplified from the unordered working-draft BAC sequencesand hybridized using a formamide-based buffer, according to theprotocol provided by Bio-Rad for Zetaprobe membranes. The posi-tive TAC clones then were ordered into a TAC contig based on theirhybridization signals.

Although eight TAC clones were transformed into cvp1-1, seedswere obtained from only five TAC clones because of low transforma-tion frequency. Restriction analysis and gene annotation indicatedthe presence of six unique genes in a complementing TAC relative toan overlapping noncomplementing TAC. Upon sequencing cvp1-1alleles of these genes, a mutation was identified only in the SMT2gene. The remaining cvp1 alleles were amplified by PCR, cloned intopCR2.1-TOPO vector (Invitrogen, Carlsbad, CA), and sequenced toidentify base-pair changes within SMT2.

Plasmid Construction and Plant Transformation

All genes used in this study were PCR-amplified products from Ara-bidopsis ecotype Columbia that were introduced into pCR2.1-TOPOcloning vector, sequenced to ensure that there were no base-pairchanges, and subcloned into the appropriate vector using compati-ble restriction sites within the polylinker. For complementation ofcvp1, a genomic clone containing the 1.5-kb upstream region, cod-ing region, and 3 untranslated region using primers MT10 (5-CGAGTGAGTCAGTCATAT-3) and MT9 (5-GTCAGTAGTGTTACT-CACACAGGC-3) was cloned into pCAMBIA 2300. For GUS con-structs, 1.5-kb upstream regions of SMT2 (MT10 and MT7 [5-CGT-

TAAGAGTGAGGAAGACC-3]) or SMT3 (SMT3-1 [5-CACAGG-GAGAAAGAGAGAAGC-3] and SMT3-2 [5-CGTGTGAGCAAATAG-ATCACG-3]) were cloned into both pBI101 (Clontech, Palo Alto, CA)and pCAMBIA 1381.

For overexpression of SMT2 or SMT3, because there are no in-trons, a genomic region representing the cDNA of SMT2 (MT1 [5-GGTCTTCCTCACTCTTAACG-3] and MT9) or SMT3 (SMT3-6 [5-CAGAGTCGTGAACTTAACG-3] and SMT3-7 [5-CCAATAGAATTT-CCCGGC-3]) was amplified by PCR and eventually cloned intopZP35 (Hajdukiewicz et al., 1994) that had been engineered with anopaline synthase 3 terminator sequence. SMT3 cDNA also was in-troduced in the antisense orientation in pZP35nos. Additional infor-mation on the design of these constructs will be provided upon re-quest. For each construct, 25 to 60 transgenic plants were analyzedfor GUS histochemistry or phenotype and/or RNA gel blot analysis.

Plant transformations were conducted (Bechtold and Pelletier,1998) with modifications (Clough and Bent, 1998).

Hormone Induction Studies and RNA Gel Blot Analysis

For hormone induction experiments, a modification of a method de-scribed previously (Gil et al., 1994) was used. One-week-old light-grown seedlings were cut into 0.5-cm pieces and incubated in 0.5 �MS salts, 1% Suc, and 50 �M cycloheximide. After 4 h of incubationat room temperature, the buffer was replaced with fresh buffer con-taining the appropriate hormone and incubated for an additional 2 h.The tissue then was blotted dry with tissue paper, frozen in liquid N2,and stored at 70�C. Hormone levels were as follows: 20 �M in-doleacetic acid, 20 �M benzyladenine, 50 �M gibberellic acid, 20�M 1-aminocyclopropane-1-carboxylic acid, and 2 �M 24-epibrass-inolide. Indoleacetic acid, gibberellic acid, 1-aminocyclopropane-1-carboxylic acid, and benzyladenine were obtained from Sigma.24-Epibrassinolide was obtained from CIDtech Research (Cambridge,Ontario, Canada). The experiment was conducted two times andyielded the same results.

For RNA gel blot analysis, RNA was isolated using the Trizolmethod (Gibco BRL). Unless noted otherwise, 10 �g of RNA wasloaded onto formaldehyde gels, electrophoresed under denaturingconditions, and transferred to a nylon membrane (Zetaprobe; Bio-Rad). To ensure equal loading of RNA, filters were stained in 0.02%methylene blue and 0.3 M NaOAc, pH 5.5, according to a publishedmethod (Herrin, 1988). After cross-linking RNA to filter, the filters werehybridized to gene-specific probes (SMT2, MT1 and MT9; SMT3,SMT3-6 and SMT3-7; SMT1, SMT1-1 [5-CTCCGATTCATCTTTATC-CTC-3] and SMT1-2 [5-GGGCATGTGCACATGATTCAG-3]) using aformamide-based hybridization buffer according to the manufac-turer’s protocol (Bio-Rad). After hybridization, filters were washed in0.5 � wash buffer (0.075 M NaCl, 0.0075 M Na-citrate, 2.5% SDS,and 0.05% NaPPi) at 55�C. Under these conditions, there was nocross-hybridization between probes. SMT1 was amplified from 1-week-old seedling RNA using the 5 rapid amplification of cDNA ends sys-tem (Gibco BRL) for first-strand synthesis. Density measurementswere made using Image Gauge version 3.3 (Fuji Photo Film Co., To-kyo, Japan).

RNA in Situ Hybridization

A full-length SMT2 probe and the SMT2 3 untranslated region (MT11[5-GGTAGAAAGGAAACATCACCGG-3] and MT9) were cloned into

Sterol Modification Alters Vein Patterns 2057

pBluescript II SK� (Stratagene) in both orientations and digestedwith the appropriate enzyme (BamHI) to yield templates for sense andantisense in vitro transcription according to the manufacturer’s sug-gestions (Roche, Indianapolis, IN). In situ hybridization experimentswere conducted (Long et al., 1996) with a hybridization temperatureof 42�C. The SMT2 3 untranslated region probe yielded the samesignal as the full-length probe and did not hybridize with SMT3. Thesignal was apparent after 2 days. Sense controls showed no signal.Photography was conducted using a Zeiss Axiophot microscope.

Measurement of Sterol Levels

Levels of 22 sterol intermediates were assayed from 1-week-oldseedlings. Plant tissue was frozen and lyophilized before extraction,purification, and gas chromatography–mass spectrometry analysis(Noguchi et al., 1999).

Upon request, all novel materials described in this article will bemade available in a timely manner for noncommercial research pur-poses. No restrictions or conditions will be placed on the use of anymaterials described in this article that would limit their use for non-commercial research purposes.

Accession Number

Accession numbers for pCAMBIA 2300 and 1381 plasmids areAF234315 and AF234302, respectively.

ACKNOWLEDGMENTS

We are grateful to Brian Keith (University of Pennsylvania) for provid-ing the cvp1 alleles and to Asami Tadao (Institute of Physical andChemical Research) for providing Brz. We thank Lien Lai (Ohio StateUniversity) for advice on positional cloning. We gratefully acknowl-edge Neil McHale (Connecticut Agricultural Experiment Station) for acritical review of the manuscript. We acknowledge the ABRC for pro-viding the TAC filter and BAC and TAC clones used in the cloning ofCVP1. This research was supported by Grants IBN-9808295 andIBN-0110730 from the National Science Foundation.

Received April 18, 2002; accepted May 27, 2002.

REFERENCES

Asami, T., and Yoshida, S. (1999). Brassinosteroid biosynthesisinhibitors. Trends Plant Sci. 4, 348–353.

Bechtold, N., and Pelletier, G. (1998). In planta Agrobacterium-mediated transformation of adult Arabidopsis thaliana plants byvacuum infiltration. Methods Mol. Biol. 82, 259–266.

Benveniste, P. (1986). Sterol biosynthesis. Annu. Rev. Plant Physiol.37, 275–308.

Bouvier-Nave, P., Husselstein, T., Desprez, T., and Benveniste,P. (1997). Identification of cDNAs encoding sterol methyl-trans-ferases involved in the second methylation step of plant sterolbiosynthesis. Eur. J. Biochem. 246, 518–529.

Busch, M., Mayer, U., and Jurgens, G. (1996). Molecular analysisof the Arabidopsis pattern formation of gene GNOM: Gene struc-ture and intragenic complementation. Mol. Gen. Genet. 250,681–691.

Carland, F.M., Berg, B.L., FitzGerald, J.N., Jinamornphongs, S.,Nelson, T., and Keith, B. (1999). Genetic regulation of vasculartissue patterning in Arabidopsis. Plant Cell 11, 2123–2137.

Carland, F.M., and McHale, N.A. (1996). LOP1: A gene involved inauxin transport and vascular patterning in Arabidopsis. Develop-ment 122, 1811–1819.

Choe, S., Dilkes, B.P., Gregory, B.D., Ross, A.S., Yuan, H., Noguchi,T., Fujioka, S., Takatsuto, S., Tanaka, A., Yoshida, S., Tax, F.E.,and Feldmann, K.A. (1999a). The Arabidopsis dwarf1 mutantis defective in the conversion of 24-methylenecholesterol tocampesterol in brassinosteroid biosynthesis. Plant Physiol. 119,897–907.

Choe, S., Noguchi, T., Fujioka, S., Takatsuto, S., Tissier, C.P.,Gregory, B.D., Ross, A.S., Tanaka, A., Yoshida, S., Tax, F.E.,and Feldmann, K.A. (1999b). The Arabidopsis dwf7/ste1 mutantis defective in the delta7 sterol C-5 desaturation step leading tobrassinosteroid biosynthesis. Plant Cell 11, 207–221.

Choe, S., Tanaka, A., Noguchi, T., Fujioka, S., Takatsuto, S.,Ross, A.S., Tax, F.E., Yoshida, S., and Feldmann, K.A. (2000).Lesions in the sterol delta reductase gene of Arabidopsis causedwarfism due to a block in brassinosteroid biosynthesis. Plant J.21, 431–443.

Clough, S.J., and Bent, A.F. (1998). Floral dip: A simplified methodfor Agrobacterium-mediated transformation of Arabidopsis thaliana.Plant J. 16, 735–743.

Clouse, S.D., and Sasse, J.M. (1998). Brassinosteroids: Essentialregulators of plant growth and development. Annu. Rev. PlantPhysiol. Plant Mol. Biol. 49, 427–451.

Cnops, G., Wang, X., Linstead, P., Van Montagu, M., VanLijsebettens, M., and Dolan, L. (2000). Tornado1 and tornado2are required for the specification of radial and circumferential pat-tern in the Arabidopsis root. Development 127, 3385–3394.

Deyholos, M.K., Cordner, G., Beebe, D., and Sieburth, L.E.(2000). The SCARFACE gene is required for cotyledon and leafvein patterning. Development 127, 3205–3213.

Diener, A.C., Li, H., Zhou, W., Whoriskey, W.J., Nes, W.D., andFink, G.R. (2000). Sterol methyltransferase 1 controls the level ofcholesterol in plants. Plant Cell 12, 853–870.

Galweiler, L., Guan, C., Muller, A., Wisman, E., Mendgen, K.,Yephremov, A., and Palme, K. (1998). Regulation of polar auxintransport by AtPIN1 in Arabidopsis vascular tissue. Science 282,2226–2230.

Geldner, N., Friml, J., Stierhof, Y.D., Jurgens, G., and Palme, K.(2001). Auxin transport inhibitors block PIN1 cycling and vesicletrafficking. Nature 413, 425–428.

Gil, P., Liu, Y., Orbovic, V., Verkamp, E., Poff, K.L., and Green,P.J. (1994). Characterization of the auxin-inducible SAUR-AC1gene for use as a molecular genetic tool in Arabidopsis. PlantPhysiol. 104, 777–784.

Grebe, M., Gadea, J., Steinmann, T., Kientz, M., Rahfeld, J.U.,Salchert, K., Koncz, C., and Jurgens, G. (2000). A conserveddomain of the Arabidopsis GNOM protein mediates subunit inter-action and cyclophilin 5 binding. Plant Cell 12, 343–356.

Hajdukiewicz, P., Svab, Z., and Maliga, P. (1994). The small, ver-satile pZP family of Agrobacterium binary vectors for plant trans-formation. Plant Mol. Biol. 25, 989–994.

Hardtke, C.S., and Berleth, T. (1998). The Arabidopsis gene

2058 The Plant Cell

MONOPTEROS encodes a transcription factor mediating embryoaxis formation and vascular development. EMBO J. 17, 1405–1411.

Hartmann, M.-A. (1998). Plant sterols and the membrane environ-ment. Trends Plant Sci. 3, 170–175.

Herrin, D.L. (1988). Rapid, reversible staining of northern blots priorto hybridization. Biotechniques 6, 196–198.

Husselstein, T., Gachotte, D., Desprez, T., Bard, M., andBenveniste, P. (1996). Transformation of Saccharomyces cerevisiaewith a cDNA encoding a sterol C-methyltransferase from Arabi-dopsis thaliana results in the synthesis of 24-ethyl sterols. FEBSLett. 381, 87–92.

Jang, J.C., Fujioka, S., Tasaka, M., Seto, H., Takatsuto, S., Ishii,A., Aida, M., Yoshida, S., and Sheen, J. (2000). A critical role ofsterols in embryonic patterning and meristem programmingrevealed by the fackel mutants of Arabidopsis thaliana. GenesDev. 14, 1485–1497.

Klahre, U., Noguchi, T., Fujioka, S., Takatsuto, S., Yokota, T.,Nomura, T., Yoshida, S., and Chua, N.H. (1998). The Arabidop-sis DIMINUTO/DWARF1 gene encodes a protein involved in ste-roid synthesis. Plant Cell 10, 1677–1690.

Koizumi, K., Sugiyama, M., and Fukuda, H. (2000). A series ofnovel mutants of Arabidopsis thaliana that are defective in the for-mation of continuous vascular network: Calling the auxin signalflow canalization hypothesis into question. Development 127,3197–3204.

Li, J., and Chory, J. (1997). A putative leucine-rich repeat receptorkinase involved in brassinosteroid signal transduction. Cell 90,929–938.

Liu, Y.G., Shirano, Y., Fukaki, H., Yanai, Y., Tasaka, M., Tabata,S., and Shibata, D. (1999). Complementation of plant mutantswith large genomic DNA fragments by a transformation-compe-tent artificial chromosome vector accelerates positional cloning.Proc. Natl. Acad. Sci. USA 96, 6535–6540.

Long, J.A., Moan, E.I., Medford, J.I., and Barton, M.K. (1996). Amember of the KNOTTED class of homeodomain proteinsencoded by the STM gene of Arabidopsis. Nature 379, 66–69.

Malamy, J.E., and Benfey, P.N. (1997). Organization and cell differ-entiation in lateral roots of Arabidopsis thaliana. Development 124,33–44.

Mayer, U., Büttner, G., and Jurgens, G. (1993). Apical-basal pat-tern formation in the Arabidopsis embryo: Studies on the role ofthe gnom gene. Development 117, 149–162.

McConnell, J.R., Emery, J., Eshed, Y., Bao, N., Bowman, J., andBarton, M.K. (2001). Role of PHABULOSA and PHAVOLUTA indetermining radial patterning in shoots. Nature 411, 709–713.

Murashige, T., and Skoog, F. (1962). A revised medium for rapidgrowth and bioassays with tobacco tissue culture. Physiol. Plant.15, 473–497.

Nelson, T., and Dengler, N. (1997). Leaf vascular pattern formation.Plant Cell 9, 1121–1135.

Nes, W.D., McCourt, B.S., Marshall, J.A., Ma, J., Dennis, A.L.,Lopez, M., Li, H., and He, L. (1999). Site-directed mutagenesis ofthe sterol methyltransferase active site from Saccharomyces cere-visiae results in formation of novel 24-ethyl sterols. J. Org. Chem.64, 1535–1542.

Noguchi, T., Fujioka, S., Takatsuto, S., Sakurai, A., Yoshida, S.,Li, J., and Chory, J. (1999). Arabidopsis det2 is defective in theconversion of (24R)-24-methylcholest-4-en-3-one to (24R)-24-

methyl-5alpha-cholestan-3-one in brassinosteroid biosynthesis.Plant Physiol. 120, 833–840.

Parks, L.W., Crowley, J.H., Leak, F.W., Smith, S.J., and Tomeo,M.E. (1999). Use of sterol mutants as probes for sterol functionsin the yeast, Saccharomyces cerevisiae. Crit. Rev. Biochem. Mol.Biol. 34, 399–404.

Peng, L., Kawagoe, Y., Hogan, P., and Delmer, D. (2002). Sito-sterol-beta-glucoside as primer for cellulose synthesis in plants.Science 295, 147–150.

Przemeck, G.K., Mattsson, J., Hardtke, C.S., Sung, Z.R., andBerleth, T. (1996). Studies on the role of the Arabidopsis geneMONOPTEROS in vascular development and plant cell axializa-tion. Planta 200, 229–237.

Ratcliffe, O.J., Riechmann, J.L., and Zhang, J.Z. (2000). INTER-FASCICULAR FIBERLESS1 is the same gene as REVOLUTA.Plant Cell 12, 315–317.

Schaeffer, A., Bronner, R., Benveniste, P., and Schaller, H.(2001). The ratio of campesterol to sitosterol that modulatesgrowth in Arabidopsis is controlled by STEROL METHYLTRANS-FERASE 2;1. Plant J. 25, 605–615.

Schindelman, G., Morikami, A., Jung, J., Baskin, T.I., Carpita,N.C., Derbyshire, P., McCann, M.C., and Benfey, P.N. (2001).COBRA encodes a putative GPI-anchored protein, which ispolarly localized and necessary for oriented cell expansion in Ara-bidopsis. Genes Dev. 15, 1115–1127.

Schrick, K., Mayer, U., Horrichs, A., Kuhnt, C., Bellini, C., Dangl,J., Schmidt, J., and Jurgens, G. (2000). FACKEL is a sterol C-14reductase required for organized cell division and expansion inArabidopsis embryogenesis. Genes Dev. 14, 1471–1484.

Sekimata, K., Kimura, T., Kaneko, I., Nakano, T., Yoneyama, K.,Takeuchi, Y., Yoshida, S., and Asami, T. (2001). A specific brassi-nosteroid biosynthesis inhibitor, Brz2001: Evaluation of its effectson Arabidopsis, cress, tobacco, and rice. Planta 213, 716–721.

Sessa, G., Steindler, C., Morelli, G., and Ruberti, I. (1998). TheArabidopsis Athb-8, -9 and -14 genes are members of a smallgene family coding for highly related HD-ZIP proteins. Plant Mol.Biol. 38, 609–622.

Shevell, D.E., Leu, W.M., Gillmor, C.S., Xia, G., Feldmann, K.A.,and Chua, N.H. (1994). EMB30 is essential for normal cell divi-sion, cell expansion, and cell adhesion in Arabidopsis andencodes a protein that has similarity to Sec7. Cell 77, 1051–1062.

Simons, K., and Ikonen, E. (1997). Functional rafts in cell mem-branes. Nature 387, 569–572.

Simons, K., and Toomre, D. (2000). Lipid rafts and signal transduc-tion. Nat. Rev. Mol. Cell Biol. 1, 31–39.

Sitbon, F., and Jonsson, L. (2001). Sterol composition and growthof transgenic tobacco plants expressing type-1 and type-2 sterolmethyltransferases. Planta 212, 568–572.

Steinmann, T., Geldner, N., Grebe, M., Mangold, S., Jackson,C.L., Paris, S., Galweiler, L., Palme, K., and Jurgens, G. (1999).Coordinated polar localization of auxin efflux carrier PIN1 byGNOM ARF GEF. Science 286, 316–318.

Wang, Z.Y., Seto, H., Fujioka, S., Yoshida, S., and Chory, J.(2001). BRI1 is a critical component of a plasma-membranereceptor for plant steroids. Nature 410, 380–383.

Zhong, R., and Ye, Z.H. (1999). IFL1, a gene regulating interfascicu-lar fiber differentiation in Arabidopsis, encodes a homeodomain-leucine zipper protein. Plant Cell 11, 2139–2152.

DOI 10.1105/tpc.003939 2002;14;2045-2058Plant Cell

Francine M. Carland, Shozo Fujioka, Suguru Takatsuto, Shigeo Yoshida and Timothy Nelson Reveals a Role for Sterols in Vascular PatterningCVP1The Identification of

 This information is current as of May 15, 2021

References /content/14/9/2045.full.html#ref-list-1

This article cites 53 articles, 25 of which can be accessed free at:

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists